How to Use Kimi K2 Thinking in Muse
Kimi K2 Thinking is built for research, coding, and writing that require multiple reasoning steps. Unlike Kimi K2 Instruct, it combines extended reasoning with tool calls to work through complex tasks.
Prompts to try with Kimi K2 Thinking in Muse
Copy an example into Muse, or ask in your own words.
Use Kimi K2 Thinking to compare PostgreSQL and SQLite for a single-server application with 20 internal users. Explain the tradeoffs, identify missing requirements, and give a recommendation in 500 words.
Use Kimi K2 Thinking to investigate whether a four-day workweek improves productivity. If research tools are available, check primary studies; distinguish measured results from claims and cite the evidence.
Use Kimi K2 Thinking to prove that the square root of 2 is irrational. State the assumptions, give a concise proof, and explain where a common incorrect argument fails.
Use Kimi K2 Thinking to review the argument in our conversation. Identify unsupported assumptions, construct the strongest counterargument, and explain what evidence would resolve the disagreement.
Use Kimi K2 Thinking to write an 800-word explanation of how database indexes affect reads and writes. Use concrete examples, explain tradeoffs, and finish with five questions that test understanding.
How to set up Kimi K2 Thinking in Muse
Pick the way that suits you. Each takes about two minutes.
Use the Muse connector
- Create an API key
Create one under API keys in the Spiral console.
- Send Muse this message
Muse reads Spiral's skill and asks for an API key.
set up https://spiral.ai/SKILL.md
- Give Muse the key
Paste the key when Muse asks for it, never in the chat itself.
Use the Spiral skill
- Send Muse this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run Kimi K2 Thinking
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openrouter", "model": "moonshotai/kimi-k2-thinking", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Muse finds Kimi K2 Thinking on Spiral, shows the price and runs it.
Kimi K2 Thinking providers and prices
Spiral runs Kimi K2 Thinking through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.01334 per run at the default inputs. $0.78 per 1M input tokens, $3.25 per 1M output tokens.
Settings you can ask for
Ask Muse for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Max tokens | 1 to 98304 | The most output tokens to generate. | |
| Temperature | 0 to 2 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 |
Every API field of MoonshotAI: Kimi K2 Thinking is on its tool page: open its card above.
What can you do with Kimi K2 Thinking in Muse?
Kimi K2 Thinking can help work through difficult questions and turn evidence into a clear explanation. State the decision you need to make, the assumptions it should examine, and the length of the answer.
Investigate a software bug
Use Kimi K2 Thinking to trace a failure through relevant code, propose a focused fix, and check it against a reproducible test.
Plan a code migration
Ask Kimi K2 Thinking to map dependencies, identify behavior that must survive the migration, and propose an ordered implementation and verification plan.
Research a difficult question
Give Kimi K2 Thinking a precise question and ask it to investigate primary sources using available research tools, resolve conflicting claims, and cite its conclusions.
Compare technical approaches
Have Kimi K2 Thinking compare designs against explicit constraints, explain tradeoffs, and identify which unanswered questions could change the recommendation.
Check a mathematical argument
Ask Kimi K2 Thinking to examine assumptions, look for counterexamples, and verify numerical steps with an available calculation tool.
Draft an analytical report
Supply source text and an audience brief, then ask Kimi K2 Thinking to build an argument, support each conclusion, and flag gaps in the evidence.
Review a terminal workflow
Use Kimi K2 Thinking to investigate a failed build or command sequence, test an explanation, and document the commands and results that support it.
Turn research into a recurring brief
Ask Kimi K2 Thinking to compare newly retrieved information with a previous summary and explain which changes affect a specific decision.
More on Kimi K2 Thinking and Muse
Overview
Kimi K2 Thinking in other agents
Questions about Kimi K2 Thinking in Muse
Can Muse use Kimi K2 Thinking?
Yes. Add Spiral to Muse as an MCP server, and Muse can find and run Kimi K2 Thinking like any built-in tool.
How much does Kimi K2 Thinking cost in Muse?
Up to $0.01334 per run at the default inputs. $0.78 per 1M input tokens, $3.25 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Muse plan is separate.
Do I need an OpenRouter account to use Kimi K2 Thinking in Muse?
You need no OpenRouter account. Muse uses one Spiral API key for every tool, and Spiral runs Kimi K2 Thinking on its own OpenRouter account.
How long does it take to set up Kimi K2 Thinking in Muse?
About two minutes: create an API key, then send Muse this message, then give Muse the key.
Does Muse wait while Kimi K2 Thinking runs?
Kimi K2 Thinking runs in the background. Muse starts the run, checks on it and gives you the text, json when it's ready.
What is Kimi K2 Thinking good at?
It is designed for research, coding, and writing that involve several reasoning steps and tool calls. Try tasks with a clear outcome, such as investigating conflicting sources, diagnosing a bug, or developing an evidence-backed report.
How much context can Kimi K2 Thinking handle?
Its published context window is 256K tokens. When preparing a long task, leave room for reasoning, tool results, and the final answer.
Can Kimi K2 Thinking analyze images or video?
It accepts text and tool definitions rather than images, audio, or video directly. Supply a transcript, extracted text, or a written description for analysis.
Why can a short answer need a large generation budget?
Reasoning and the final answer draw from the same generation budget. Set the desired answer length in your prompt, and leave enough generation room for the reasoning needed to produce it.
What temperature should I use?
Moonshot recommends 1.0 for Kimi K2 Thinking. Start there, and assess results using representative tasks with clear checks for correctness.
Start using Kimi K2 Thinking in Muse
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Muse:
set up https://spiral.ai/SKILL.md


